Simulating boreal forest carbon dynamics after stand-replacing fire disturbance: insights from a global process-based vegetation model

摘要

Stand-replacing fires are the dominant fire type in North American borealforests. They leave a historical legacy of a mosaic landscape of differentaged forest cohorts. This forest age dynamics must be included in vegetationmodels to accurately quantify the role of fire in the historical and currentregional forest carbon balance. The present study adapted the globalprocess-based vegetation model ORCHIDEE to simulate the CO₂ emissionsfrom boreal forest fire and the subsequent recovery after a stand-replacingfire; the model represents postfire new cohort establishment, forest standstructure and the self-thinning process. Simulation results are evaluatedagainst observations of three clusters of postfire forest chronosequences inCanada and Alaska. The variables evaluated include: fire carbon emissions,CO₂ fluxes (gross primary production, total ecosystem respiration andnet ecosystem exchange), leaf area index, and biometric measurements(aboveground biomass carbon, forest floor carbon, woody debris carbon, standindividual density, stand basal area, and mean diameter at breast height).When forced by local climate and the atmospheric CO₂ history at eachchronosequence site, the model simulations generally match the observedCO₂ fluxes and carbon stock data well, with model-measurement meansquare root of deviation comparable with the measurement accuracy (forCO₂ flux ~100 g C m?2 yr?1, for biomass carbon~1000 g C m?2 and for soil carbon ~2000 g C m−2). We find that the current postfire forest carbon sink at theevaluation sites, as observed by chronosequence methods, is mainly due to acombination of historical CO₂ increase and forest succession. Climatechange and variability during this period offsets some of these expectedcarbon gains. The negative impacts of climate were a likely consequence ofincreasing water stress caused by significant temperature increases thatwere not matched by concurrent increases in precipitation. Our simulationresults demonstrate that a global vegetation model such as ORCHIDEE is ableto capture the essential ecosystem processes in fire-disturbed borealforests and produces satisfactory results in terms of both carbon fluxes andcarbon-stock evolution after fire. This makes the model suitable forregional simulations in boreal regions where fire regimes play a key role inthe ecosystem carbon balance.